Description
Audit dimension 4 confirmed that the shipped resource-sizing contract still outruns its evidence. charts/meraki-dashboard-exporter/values.yaml:384-412, the chart README and docs/scaling-guide.md:232-255 prescribe small, medium and large CPU and memory quantities. The same guide at lines 257-265 states that the largest actual measurement is a one-network HOMELAB replay, not a medium or dense-switch measurement; the retained fixture has 19 devices. MDE-0002 remains permanently Parked because no reachable organization has the required medium or dense-switch topology. This task must not invent replacement numbers or move defaults: unsupported recommendations should be removed or labeled non-authoritative until representative evidence exists.
Acceptance Criteria
- #1 Every published resource quantity is either the unchanged bootable chart default or is paired with a representative cited measurement and stated assumptions
- #2 Unsupported medium and large resource recommendations are removed or explicitly marked non-authoritative without replacement estimates
- #3 Chart values, chart README and scaling guidance use identical evidence labels and retain the measured HOMELAB result as HOMELAB only
Definition of Done
- #1 just check (ruff format –check, ruff check, mypy, generated-doc drift, offline API conformance, and the marker-filtered pytest run with the 80% coverage floor — this is exactly what the CI
testjob runs) - #2 just gen, when metrics, config, endpoints, collectors, the settings schema or the chart config changed —
just checkincludes the drift gate and CI fails the build on it - #3 Grafana queries in grafana/dashboards/*.json and grafana/alerts/ updated, if a metric or label name changed
Implementation Plan
Remove or mark non-authoritative every unsupported medium/large resource quantity across values, chart README and scaling guidance; retain the bootable default and the HOMELAB measurement without inventing replacements.
Final Summary
Removed unsupported resource-tier recommendations and retained only measured operational guidance. Documentation validation, just check, and just ci passed at 5e8d9c23b76a2f2edd531c15c776cbfbcc9134fa; exact-head CI 33843833956 and publication 33843966820 succeeded.